基于模糊系统的高超声速飞行器鲁棒控制方法研究

CASIA OpenIR > 毕业生 > 博士学位论文

	基于模糊系统的高超声速飞行器鲁棒控制方法研究
	刘轶凡1,2
	2018-05-30
学位类型	工学博士
英文摘要	高超声速飞行器是指飞行马赫数大于5的有翼或无翼飞行器。与现有航空器及航天器相比，高超声速飞行器具有飞行速度快、飞行高度高、快速全球到达、突防能力强、作战效能高等优点，具有巨大的军事和民用价值，因而受到全球研究者的重点关注。在实际高速飞行过程中，由于发动机/机身一体化设计、机体模态造成的不可忽视的弹性效应及所处的恶劣空天环境造成该飞行器动力学容易受到外界环境的强干扰和测量噪声的影响，并且具有强耦合、强非线性、强不确定性、快时变特性等特性，这对控制系统的设计要求带来巨大的挑战。模糊系统及其控制方法作为处理非线性不确定系统的有效方法被广泛地应用于飞行器控制中。因此，本论文以高超声速飞行器纵向模型为研究对象，基于模糊系统理论，针对鲁棒飞行器控制系统关键问题进行深入研究，探讨模糊系统在高超声速飞行器控制中的应用潜力。本课题即将开展的工作包括以下几个部分：（1）建立高超声速飞行器的纵向模型，以及对气动力/力矩方程和执行器等动态环节建模。推导面向控制的高超声速飞行器模型，为后续控制系统设计奠定了基础。（2）针对高超声速飞行器飞行过程中受到不确定性干扰问题，提出基于LPV T-S模糊系统的高超声速飞行器鲁棒控制方法。通过分析高超声速飞行器模型中主要影响飞行稳定性的主导不确定参数，建立含参数不确定性的高超声速飞行器动力学模型。并在此基础上，建立飞行器LPV T-S模糊模型。设计模糊鲁棒跟踪反馈控制器，保证闭环系统的极点配置到一个期望的复平面区域，实现对期望速度和高度的跟踪。（3）针对高超声速飞行器仅部分状态可测问题，提出基于观测器的二型模糊自适应鲁棒输出反馈控制方法。该控制方法首先引入二型模糊函数估计器，对动力学模型中的未知函数进行估计。然后，结合二型模糊函数估计器，设计降阶状态观测器对不可测状态实现实时观测。同时，为避免出现传统反步法中的“微分项膨胀”问题，设计指令滤波反步控制器，实现对高度速度的跟踪控制。最后综合状态观测器和反步控制器，构成高超声速飞行器速度和高度跟踪控制系统，并实现其稳定性的证明。（4）针对高超声速飞行器姿态子系统控制器受测量噪声的影响问题，提出期望补偿模糊自适应控制方法。首先设计一型模糊估计器，对系统中未知函数进行估计。然后，结合期望补偿自适应技术，用未受噪声信号影响的期望指令值替换模糊估计器自适应律中的状态测量值，设计出期望补偿模糊自适应控制器。该方法可以避免噪声信号对自适应律的直接影响，并保证姿态子系统在噪声环境下的有效跟踪控制。（5）针对高超声速飞行器纵向模型全状态均受测量噪声影响下的控制问题，提出一种基于模糊滤波的自适应鲁棒控制方法。对含测量噪声的状态信号，首先利用模糊状态滤波器对测量状态进行滤波处理，得到滤除噪声后的状态信号。然后，对高度子系统和速度子系统设计基于动态面反步法的模糊自适应控制器，以进一步削弱噪声对虚拟控制量和最终控制器输出的影响。最后结合模糊滤波器和控制器构成完整的闭环控制系统，并对其稳定性进行了分析。以上各控制方法均通过仿真实验验证其有效性。综上所述，本论文深入研究了基于模糊系统理论的高超声速飞行器鲁棒模糊控制方法的若干关键问题，验证了飞行器分别在强不确定性、仅部分状态可测和测量状态受噪声污染条件影响下的控制系统鲁棒性，为模糊系统理论在高超声速飞行器控制中的应用挖掘出潜力。 ; Hypersonic vehicles refer to the winged or wingless aircrafts with velocity more than Mach 5. Compared with the existing aircraft and spacecraft, hypersonic vehicle has faster flight speed, higher flight altitude, and the ability of rapid global arrival and strong penetration. Hence, it has a great value in military and civil applications which attracts attention of global researchers. During the high-speed flight, the elastic effect caused by the engine/airframe integration design, the disturbance and measurement noises caused by the poor space environment and the strong coupling, nonlinearity, uncertainty, time-varying characteristics caused by flight dynamics bring a great challenge in control system design. Fuzzy system based control approachs act as effecitive ways for solving the control problem of nonlinear uncertain systems, which have been widely used in aircraft control. Therefore, based on the fuzzy logic system, the control problems for hypersonic vehicles are investigated in this paper in order to exploring the potential of the fuzzy logic system based control approachs. This paper is to carry out the work including the following parts: (1) The longitudinal model of hypersonic vehicle is constructed. And the aerodynamic force / torque equation and the actuator dynamic model are established. Then the control-oriented models of the hypersonic vehicle are derived, which act as the foundation for the coming control system design. (2) Considering the strong uncertainty which affects the stability of the hypersonic vehicle, an LPV T-S fuzzy controller is proposed for the robust control of the hypersonic vehicle. Through the analysis of the main uncertain parameters which affect the stability of the hypersonic vehicle, a hypersonic vehicle dynamics model with parametric uncertainties is constructed. On this basis, the LPV T-S fuzzy model of the vehicle is established. Then fuzzy robust tracking feedback controller is designed which can ensure the poles of the closed-loop system located in the desired complex plane area. Hence the proposed control approach can achieve good performance of the velocity and altitude tracking performances. (3) For the output feedback control problem of the hypersonic vehicle, an observer-based type-2 fuzzy adaptive robust controller is designed. The controller adopts type-2 fuzzy system for unknown function approximation. Then, by combining with the type-2 fuzzy approximation, a reduced order state observer for estiamaiton of the unmeasurable state is constructed. In addition, to avoid the problem of “explosion of complexity”in the traditional backstepping method, the command filter based backstepping controller is designed, which realizes the altitude and velocity tracking control. Finally, the stability of the closed-loop system is analyzed. (4) For the control problem of the hypersonic vehicles’ attitude subsystem affected by the measurement noises, the desired compensation adaptive fuzzy control method is proposed. At first, a type-1 fuzzy system based function approximator is designed for the estimation of the unknown function. Then, with the desired compensation adaptive techtique, the desired reference signals are used in adaptive term for replacing the measurement state signals. This approach has the advantages that it avoids the noise signal directly affecting the adaptive laws and can guarantee the tracking performance of the controller when the measurement noises exist. (5) For the control problem of the longitudinal hypersonic vehicle model when full states are subject to measurement noises, this paper proposes an adaptive type-2 fuzzy filter based control method. At first, by using the fuzzy filter, the state signals are extracted from the raw signal contaminated with noises. Next, dynamic surface backstepping based adaptive fuzzy controller is designed for the altitude and velocity subsystems, which has the adavantages reducing the effect of noises on the virtual and final control laws. Finally, the stability of the closed-loop system is analyzed. All controllers are verified by simulations and the results show their effectiveness. In summary, this thesis has studied the problems of hypersonic vehicle fuzzy robust fuzzy controller design based on the theory of fuzzy logic systems, and verified the robust fuzzy control against the strong uncertainties, the output feedback fuzzy control and the robust fuzzy control against the measurement noises. This work provides the foundation for hypersonic vehicles fuzzy control approaches.
关键词	高超声速飞行器模糊逻辑系统自适应控制鲁棒控制输出反馈测量噪声
文献类型	学位论文
条目标识符	http://ir.ia.ac.cn/handle/173211/21075
专题	毕业生_博士学位论文
作者单位	1.中国科学院自动化研究所 2.中国科学院大学
第一作者单位	中国科学院自动化研究所
推荐引用方式 GB/T 7714	刘轶凡. 基于模糊系统的高超声速飞行器鲁棒控制方法研究[D]. 北京. 中国科学院研究生院,2018.

条目包含的文件
文件名称/大小	文献类型	版本类型	开放类型	使用许可
刘轶凡-博士论文书签版-知识库上传.pd（4753KB）	学位论文		限制开放	CC BY-NC-SA